BSAN2201 Chap.3 Descriptive Analysis and Visual Choice
Descriptive Analysis and Visual Choice
Descriptive analytics compresses current or historical observations into patterns a decision-maker can understand. Measures of central tendency describe the centre, while range and standard deviation describe spread. Averages can disagree because they answer different questions, especially when a distribution is skewed or contains an extreme value.
Visual choice also follows the comparison: bars compare categories, lines show movement through time, scatterplots show relationships and composition charts show genuine parts of a whole. The chapter emphasises axes, denominators, units and aggregation so that presentation does not exaggerate the evidence.
Interpretation begins before styling.
The analyst checks the observation period, denominator, units, missing values and aggregation level, then chooses a display whose visual channel matches the comparison. Outliers are verified and tested rather than removed automatically. Subgroup patterns are considered where case mix could distort an overall average, but tiny slices are not treated as stable evidence.
A complete descriptive statement names the pattern, quantifies its size, reports variability and states what remains unexplained. That final boundary is essential because a summary can identify where to investigate without establishing why the pattern occurred.
What this chapter covers
- 01
Mean, median and mode
- 02
Range and standard deviation
- 03
Outliers and sensitivity
- 04
Category comparisons
- 05
Time and relationship charts
- 06
Axes, denominators and units
Worked example · free
Explain an extreme delivery time
- +1Calculate the mean: 180 divided by 6 equals 30 minutes.
- +1Calculate the median: the middle pair 20 and 21 gives 20.5 minutes.
- +1Use the median for the ordinary customer experience because the distribution is skewed.
- +1Retain the mean and investigate the 80-minute event because it affects workload and risk.
Key terms
- Central tendency
- Statistical descriptions of the centre of a dataset.
- Mean
- The sum of values divided by their count.
- Median
- The middle value after observations are ordered.
- Mode
- The most frequently occurring value or category.
- Standard deviation
- A measure of dispersion around the mean.
Descriptive Analysis and Visual Choice FAQ
When should I prefer the median to the mean?
Prefer the median when the distribution is strongly skewed or a few extreme observations would make the mean unrepresentative of the middle experience. Keep the mean when total burden or resource use matters, and report both when they reveal different operational facts.
How do I choose between a bar and line chart?
Use bars when comparing categories and a line when ordered time is essential to the pattern. The connected line implies sequence, so it is misleading for unrelated categories. Label units and keep the axis range appropriate to the decision.
Should an outlier always be removed?
No. First verify whether it is a recording error, a special event or part of the real process. Show how the result changes with and without it. Removing valid extreme evidence can hide the operational risk that deserves the most attention.
How can aggregation change a conclusion?
An overall average can combine groups with different sizes, baselines or trends and conceal what happens inside each group. Recalculate the measure by a relevant segment and inspect both counts and rates. If the direction changes after disaggregation, report the composition effect instead of treating the headline number as universal.
Assessment move
Take a small dataset and calculate mean, median, range and standard deviation, then explain what each adds. Redraw the same values as an honest and a misleading chart to practise axis auditing. For every visual, write one sentence on the pattern, one on its magnitude and one on the denominator or scale that could change the interpretation.
Rebuild one lecture visual from a raw table and confirm that category order, time order and totals survive. Compare counts with rates whenever group sizes differ, and label the denominator beside the calculation. Run a sensitivity check with and without a verified extreme value, then explain whether the business conclusion changes. End by writing the next question the descriptive evidence raises.
This habit prevents a chart from being mistaken for a causal answer. Create three versions of the same display for a trend, comparison and distribution, then choose the one whose visual task matches the question. Audit the zero baseline, interval width, category order and annotation before interpreting it. Split the data by one meaningful segment and check whether the aggregate direction persists.
Write a two-sentence executive interpretation that reports magnitude and uncertainty without claiming a cause. Next, sketch what additional design or evidence would be needed to investigate the mechanism. This keeps descriptive work valuable while preserving the boundary between pattern recognition and explanation.
Working through Descriptive Analysis and Visual Choice in BSAN2201? Sia is AskSia’s AI Business and Economics tutor — ask any BSAN2201 Descriptive Analysis and Visual Choice question and get a clear, step-by-step explanation grounded in how BSAN2201 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.